English

From Knowing to Doing Precisely: A General Self-Correction and Termination Framework for VLA models

Robotics 2026-02-03 v1

Abstract

While vision-language-action (VLA) models for embodied agents integrate perception, reasoning, and control, they remain constrained by two critical weaknesses: first, during grasping tasks, the action tokens generated by the language model often exhibit subtle spatial deviations from the target object, resulting in grasp failures; second, they lack the ability to reliably recognize task completion, which leads to redundant actions and frequent timeout errors. To address these challenges and enhance robustness, we propose a lightweight, training-free framework, VLA-SCT. This framework operates as a self-correcting control loop, combining data-driven action refinement with conditional logic for termination. Consequently, compared to baseline approaches, our method achieves consistent improvements across all datasets in the LIBERO benchmark, significantly increasing the success rate of fine manipulation tasks and ensuring accurate task completion, thereby promoting the deployment of more reliable VLA agents in complex, unstructured environments.

Keywords

Cite

@article{arxiv.2602.01811,
  title  = {From Knowing to Doing Precisely: A General Self-Correction and Termination Framework for VLA models},
  author = {Wentao Zhang and Aolan Sun and Wentao Mo and Xiaoyang Qu and Yuxin Zheng and Jianzong Wang},
  journal= {arXiv preprint arXiv:2602.01811},
  year   = {2026}
}

Comments

Accepted to 2026 IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP 2026)

R2 v1 2026-07-01T09:31:17.226Z